We present a quantum protocol which securely and implicitly implements a random shuffle to realize differential privacy in the shuffle model. The shuffle model of differential privacy amplifies privacy achievable via local differential privacy by randomly permuting the tuple of outcomes from data contributors. In practice, one needs to address how this shuffle is implemented. Examples include implementing the shuffle via mix-networks, or shuffling via a trusted third party. These implementation specific issues raise non-trivial computational and trust requirements in a classical system. We propose a quantum version of the protocol using entanglement of quantum states and show that the shuffle can be implemented without these extra requirements. Our protocol implements κ -ary randomized response, for any value of κ≥ 2 , and furthermore, can be efficiently implemented using fault-tolerant computation.

Large Language Models (LLMs) generate responses based on user prompts. Often, these prompts may contain highly sensitive information, including personally identifiable information (PII), which could be exposed to third parties hosting these models. In this work, we propose a new method to sanitize user prompts. Our mechanism uses the randomized response mechanism of differential privacy to randomly and independently perturb each character in a word. The perturbed text is then sent to a remote LLM, which first performs a prompt restoration and subsequently performs the intended downstream task. The idea is that the restoration will be able to reconstruct non-sensitive words even when they are perturbed due to cues from the context, as well as the fact that these words are often very common. On the other hand, perturbation would make reconstruction of sensitive words difficult because they are rare. We experimentally validate our method on two datasets, i2b2/UTHealth and Enron, using two LLMs: Llama-3.1 8B Instruct and GPT-4o mini. We also compare our approach with a word-level differentially private mechanism, and with a rule-based PII redaction baseline, using a unified privacy-utility evaluation. Our results show that sensitive PII tagged in these datasets are reconstructed at a rate close to the theoretical rate of reconstructing completely random words, whereas non-sensitive words are reconstructed at a much higher rate. Our method has the advantage that it can be applied without explicitly identifying sensitive pieces of information in the prompt, while showing a good privacy-utility tradeoff for downstream tasks.

Differential privacy is a mathematical notion of data privacy that has fast become the de facto standard in privacy-preserving data analysis. Recently a lot of work has focused on differential privacy in the quantum setting. Continuing on this line of study, we investigate how to answer counting queries on a quantum encoded dataset with differential privacy. An example of a counting query is “How many people in the dataset are over the age of 25 and with a university education?” Counting queries form the most basic but nonetheless rich set of statistics extractable from a dataset. We show that answering these queries on a quantum encoded dataset reduces to measuring the amplitude of one of two orthogonal states. We then analyze the differential privacy properties of two algorithms from literature to measure amplitude: one which performs repeated measurements in the computational basis, and the other which utilizes the classic amplitude estimation algorithm. For the first technique, we prove privacy results for the case of counting queries that improve on previously known results on general queries, and show that the mechanism in fact amplifies privacy due to inherent randomness. For the second method, we derive a tight bound on maximum possible change in the amplitude if we add or remove a single item in the dataset, a quantity called global sensitivity which is central in making an algorithm differentially private. We then show a differentially private version of the amplitude estimation algorithm for counting queries. We also discuss how these methods can be outsourced to a quantum server to blindly compute counting queries with differential privacy.

A widely used method to ensure privacy of unstructured text data is the multidimensional Laplace mechanism for d_X-privacy, which is a relaxation of differential privacy for metric spaces. We identify an intriguing peculiarity of this mechanism. When applied on a word-by-word basis, the mechanism either outputs the original word, or completely dissimilar words, and very rarely outputs semantically similar words. We investigate this observation in detail, and tie it to the fact that the distance of the nearest neighbor of a word in any word embedding model (which are high-dimensional) is much larger than the relative difference in distances to any of its two consecutive neighbors. We also show that the dot product of the multidimensional Laplace noise vector with any word embedding plays a crucial role in designating the nearest neighbor. We derive the distribution, moments and tail bounds of this dot product. We further propose a fix as a post-processing step, which satisfactorily removes the above-mentioned issue.

Software binary analysis, tools and machine learning aid security analysts in interpreting data, by automated means that filter, prioritize, and arrange pertinent information for skilled analysts. In this work, we revisit cooperative human-machine teams and evaluate the possibility of enabling untrained humans to assist machines and skilled analysts in their analysis of software binaries. Specifically, we propose a pipeline to transform a complex input domain into facial images on which untrained individuals make similarity decisions. Our faces include realistic human, animal, artistic, and anime faces that preserve inherent distances between data points of the input domain. Our approach is evaluated through a human study, where untrained respondents with minimal training successfully flag machine misclassifications. The untrained human does not replace the machine or skilled analyst, instead, utilized in a triage setting, to identify samples without historical precedence, deferring the decision to the skilled analyst for deeper inspection.

Millions of users leverage generative pretrained transformer (GPT)-based language models developed by leading model providers for a wide range of tasks. To support enhanced user interaction and customization, many platforms-such as OpenAI-now enable developers to create and publish tailored model instances, known as custom GPTs, via dedicated repositories or application stores. These custom GPTs empower users to browse and interact with specialized applications designed to meet specific needs. However, as custom GPTs see growing adoption, concerns regarding their security vulnerabilities have intensified. Existing research on these vulnerabilities remains largely theoretical, often lacking empirical, large-scale, and statistically rigorous assessments of associated risks. In this study, we analyze 14,904 custom GPTs to assess their susceptibility to seven exploitable threats, such as roleplay-based attacks, system prompt leakage, phishing content generation, and malicious code synthesis, across various categories and popularity tiers within the OpenAI marketplace. We introduce a multi-metric ranking system to examine the relationship between a custom GPT's popularity and its associated security risks. Our findings reveal that over 95 protections. The most prevalent vulnerabilities include roleplay-based vulnerabilities (96.51 (91.22 inherent security weaknesses, which are often inherited or amplified in custom GPTs. These results highlight the urgent need for enhanced security measures and stricter content moderation to ensure the safe deployment of GPT-based applications.

Millions of users leverage generative pretrained transformer (GPT)-based language models developed by leading model providers for a wide range of tasks. To support enhanced user interaction and customization, many platforms-such as OpenAI-now enable developers to create and publish tailored model instances, known as custom GPTs, via dedicated repositories or application stores. These custom GPTs empower users to browse and interact with specialized applications designed to meet specific needs. However, as custom GPTs see growing adoption, concerns regarding their security vulnerabilities have intensified. Existing research on these vulnerabilities remains largely theoretical, often lacking empirical, large-scale, and statistically rigorous assessments of associated risks. In this study, we analyze 14,904 custom GPTs to assess their susceptibility to seven exploitable threats, such as roleplay-based attacks, system prompt leakage, phishing content generation, and malicious code synthesis, across various categories and popularity tiers within the OpenAI marketplace. We introduce a multi-metric ranking system to examine the relationship between a custom GPT's popularity and its associated security risks. Our findings reveal that over 95% of custom GPTs lack adequate security protections. The most prevalent vulnerabilities include roleplay-based vulnerabilities (96.51%), system prompt leakage (92.20%), and phishing (91.22%). Furthermore, we demonstrate that OpenAI's foundational models exhibit inherent security weaknesses, which are often inherited or amplified in custom GPTs. These results highlight the urgent need for enhanced security measures and stricter content moderation to ensure the safe deployment of GPT-based applications.

Adversarial examples add imperceptible alterations to inputs with the objective to induce misclassification in machine learning models. They have been demonstrated to pose significant challenges in domains like image classification, with results showing that an adversarially perturbed image to evade detection against one classifier is most likely transferable to other classifiers. Adversarial examples have also been studied in malware analysis. Unlike images, program binaries cannot be arbitrarily perturbed without rendering them non-functional. Due to the difficulty of crafting adversarial program binaries, there is no consensus on the transferability of adversarially perturbed programs to different detectors. In this work, we explore the robustness of malware detectors against adversarially perturbed malware. We investigate the transferability of adversarial attacks developed against one detector, against other machine learning-based malware detectors with different feature space, and code similarity techniques, specifically, locality sensitive hashing-based detectors. Our analysis reveals that adversarial program binaries crafted for one detector are generally less effective against others. We also evaluate an ensemble of detectors and show that they can potentially mitigate the impact of adversarial program binaries. Finally, we demonstrate that substantial program changes made to evade detection may result in the transformation technique being identified, implying that the adversary must make minimal changes to the program binary.

Large Language Models (LLMs) are often fine-tuned to adapt their general-purpose knowledge to specific tasks and domains such as cyber threat intelligence (CTI). Fine-tuning is mostly done through proprietary datasets that may contain sensitive information. Owners expect their fine-tuned model to not inadvertently leak this information to potentially adversarial end users. Using CTI as a use case, we demonstrate that data-extraction attacks can recover sensitive information from fine-tuned models on CTI reports, underscoring the need for mitigation. Retraining the full model to eliminate this leakage is computationally expensive and impractical. We propose an alternative approach, which we call privacy alignment, inspired by safety alignment in LLMs. Just like safety alignment teaches the model to abide by safety constraints through a few examples, we enforce privacy alignment through few-shot supervision, integrating a privacy classifier and a privacy redactor, both handled by the same underlying LLM. We evaluate our system, called CTIGuardian, using GPT-4o mini and Mistral-7B Instruct models, benchmarking against Presidio, a named entity recognition (NER) baseline. Results show that CTIGuardian provides a better privacy-utility trade-off than NER based models. While we demonstrate its effectiveness on a CTI use case, the framework is generic enough to be applicable to other sensitive domains.

Increased cloud adoption in healthcare has amplified ransomware and malware threats, accounting for 19% of global breaches in 2024. Despite this surge, the behavior of attackers exploiting healthcare systems remains under-explored in academic literature. This paper bridges that gap by deploying a scalable and stealthy deception network specifically designed for healthcare environments. The network comprises 30 real-world vulnerable healthcare web applications, mimicking domain-specific workflows across multi-cloud infrastructures, such as patient registration and billing. We leveraged ATTACK-BERT to generate semantic embeddings and applied co-regularized spectral clustering with normalized cuts to analyze multi-protocol attack traffic. Our analysis revealed nuanced attacker behaviors, including regional and protocol-specific variations, exploitation of healthcare protocols like HL7, and the use of encryption to bypass detection. A comparative sub-study further showed that attackers deliberately engage with vulnerable systems, highlighting the strategic value of deception-based defenses. By focusing on behavioral insights within healthcare-specific settings, this work lays the groundwork for integrating deception into the broader security posture of critical infrastructures.

Two parties wish to collaborate on their datasets. However, before they reveal their datasets to each other, the parties want to have the guarantee that the collaboration would be fruitful. We look at this problem from the point of view of machine learning, where one party is promised an improvement on its prediction model by incorporating data from the other party. The parties would only wish to collaborate further if the updated model shows an improvement in accuracy. Before this is ascertained, the two parties would not want to disclose their models and datasets. In this work, we construct an interactive protocol for this problem based on the fully homomorphic encryption scheme over the Torus (TFHE) and label differential privacy, where the underlying machine learning model is a neural network. Label differential privacy is used to ensure that computations are not done entirely in the encrypted domain, which is a significant bottleneck for neural network training according to the current state-of-the-art FHE implementations. We formally prove the security of our scheme assuming honest-but-curious parties, but where one party may not have any expertise in labelling its initial dataset. Experiments show that we can obtain the output, i.e., the accuracy of the updated model, with time many orders of magnitude faster than a protocol using entirely FHE operations.

Users may inadvertently upload personally identifiable information (PII) to Machine Learning as a Service (MLaaS) providers. When users no longer want their PII on these services, regulations like GDPR and COPPA mandate a right to forget for these users. As such, these services seek efficient methods to remove the influence of specific data points. Thus the introduction of machine unlearning. Traditionally, unlearning is performed with the removal of entire data samples (sample unlearning) or whole features across the dataset (feature unlearning). However, these approaches are not equipped to handle the more granular and challenging task of unlearning specific objects within a sample. To address this gap, we propose a scene graph-based object unlearning framework. This framework utilizes scene graphs, rich in semantic representation, transparently translate unlearning requests into actionable steps. The result, is the preservation of the overall semantic integrity of the generated image, bar the unlearned object. Further, we manage high computational overheads with influence functions to approximate the unlearning process. For validation, we evaluate the unlearned object's fidelity in outputs under the tasks of image reconstruction and image synthesis. Our proposed framework demonstrates improved object unlearning outcomes, with the preservation of unrequested samples in contrast to sample and feature learning methods. This work addresses critical privacy issues by increasing the granularity of targeted machine unlearning through forgetting specific object-level details without sacrificing the utility of the whole data sample or dataset feature.

Individuals have been increasingly interacting with online Large Language Models (LLMs), both in their work and personal lives. These interactions raise privacy issues as the LLMs are typically hosted by third-parties who can gather a variety of sensitive information about users and their companies. Text Sanitization techniques have been proposed in the literature and can be used to sanitize user prompts before sending them to the LLM. However, sanitization has an impact on the downstream task performed by the LLM, and often to such an extent that it leads to unacceptable results for the user. This is not just a minor annoyance, with clear monetary consequences as LLM services charge on a per use basis as well as great amount of computing resources wasted. We propose an architecture leveraging a Small Language Model (SLM) at the user-side to help estimate the impact of sanitization on a prompt before it is sent to the LLM, thus preventing resource losses. Our evaluation of this architecture revealed a significant problem with text sanitization based on Differential Privacy, on which we want to draw the attention of the community for further investigation.

We present ConvoCache, a conversational caching system that solves the problem of slow and expensive generative AI models in spoken chatbots. ConvoCache finds a semantically similar prompt in the past and reuses the response. In this paper we evaluate ConvoCache on the DailyDialog dataset. We find that ConvoCache can apply a UniEval coherence threshold of 90% and respond to 89% of prompts using the cache with an average latency of 214ms, replacing LLM and voice synthesis that can take over 1s. To further reduce latency we test prefetching and find limited usefulness. Prefetching with 80% of a request leads to a 63% hit rate, and a drop in overall coherence. ConvoCache can be used with any chatbot to reduce costs by reducing usage of generative AI by up to 89%.

One-Time Passwords (OTPs) are a crucial component of multi-factor authentication (MFA) systems, providing additional security by requiring users to supply a dynamically generated code for authenticating to web services. The growth in smartphone usage has resulted in a shift from hardware tokens to mobile app-based OTP authenticators; however, these apps also present potential security and privacy threats. In this paper, we present a comprehensive analysis of 182 publicly available OTP apps on Google Play. Our analysis entails an array of passive and active measurements meticulously designed to assess the security and privacy attributes inherent to each OTP application. We investigate the presence of suspicious libraries, usage of binary protections, access to root privileges, secure backup and cryptographic mechanisms, and protection against traffic interception. Our experiments highlight several security and privacy weaknesses in instances of OTP apps. We observe that 28

Energy consumption data is collected the service providers and shared with various stakeholders involved in a smart grid ecosystem. The fine-grained energy consumption data is immensely useful for maintaining and operating grid services. Further, these data can be used for future consumption prediction using machine learning and statistical models and market segmentation purposes. However, sharing and releasing fine-grained energy data or releasing predictive models trained on user-specific data induce explicit violations of private information of consumers [34, 41]. Thus, the service providers may share and release aggregated statistics to protect the privacy of users aiming at mitigating the privacy risks of individual users’ consumption traces. In this chapter, we show that an attacker can recover individual users’ traces of energy consumption data by exploiting regularity and uniqueness properties of individual consumption load patterns. We propose an unsupervised attack framework to recover hourly energy consumption time-series of users without any background information. We construct the problem of assigning aggregated energy consumption meter readings to individual users as a mathematical assignment problem and solve it by the Hungarian algorithm [30, 50]. We used two real-world datasets to demonstrate an attacker’s performance in recovering private traits of users. Our results show that an attacker is capable of recovering 70

One of the new developments in chit-chat bots is a long-term memory mechanism that remembers information from past conversations for increasing engagement and consistency of responses. The bot is designed to extract knowledge of personal nature from their conversation partner, e.g., stating preference for a particular color. In this paper, we show that this memory mechanism can result in unintended behavior. In particular, we found that one can combine a personal statement with an informative statement that would lead the bot to remember the informative statement alongside personal knowledge in its long term memory. This means that the bot can be tricked into remembering misinformation which it would regurgitate as statements of fact when recalling information relevant to the topic of conversation. We demonstrate this vulnerability on the BlenderBot 2 framework implemented on the ParlAI platform and provide examples on the more recent and significantly larger BlenderBot 3 model. We generate 150 examples of misinformation, of which 114 (76

Named entity recognition models (NER), are widely used for identifying named entities (e.g., individuals, locations, and other information) in text documents. Machine learning based NER models are increasingly being applied in privacy-sensitive applications that need automatic and scalable identification of sensitive information to redact text for data sharing. In this paper, we study the setting when NER models are available as a black-box service for identifying sensitive information in user documents and show that these models are vulnerable to membership inference on their training datasets. With updated pre-trained NER models from spaCy, we demonstrate two distinct membership attacks on these models. Our first attack capitalizes on unintended memorization in the NER's underlying neural network, a phenomenon NNs are known to be vulnerable to. Our second attack leverages a timing side-channel to target NER models that maintain vocabularies constructed from the training data. We show that different functional paths of words within the training dataset in contrast to words not previously seen have measurable differences in execution time. Revealing membership status of training samples has clear privacy implications. For example, in text redaction, sensitive words or phrases to be found and removed, are at risk of being detected in the training dataset. Our experimental evaluation includes the redaction of both password and health data, presenting both security risks and a privacy/regulatory issues. This is exacerbated by results that indicate memorization after only a single phrase. We achieved a 70% AUC in our first attack on a text redaction use-case. We also show overwhelming success in the second timing attack with an 99.23% AUC. Finally we discuss potential mitigation approaches to realize the safe use of NER models in light of the presented privacy and security implications of membership inference attacks.

Biometric authentication service providers often claim that it is not possible to reverse-engineer a user's raw biometric sample, such as a fingerprint or a face image, from its mathematical (feature-space) representation. In this paper, we investigate this claim on the specific example of deep neural network (DNN) embeddings. Inversion of DNN embeddings has been investigated for explaining deep image representations or synthesizing normalized images. Existing studies leverage full access to all layers of the original model, as well as all possible information on the original dataset. For the biometric authentication use case, we need to investigate this under adversarial settings where an attacker has access to a feature-space representation but no direct access to the exact original dataset nor the original learned model. Instead, we assume varying degree of attacker's background knowledge about the distribution of the dataset as well as the original learned model (architecture and training process). In these cases, we show that the attacker can exploit off-the-shelf DNN models and public datasets, to mimic the behaviour of the original learned model to varying degrees of success, based only on the obtained representation and attacker's prior knowledge. We propose a two-pronged attack that first infers the original DNN by exploiting the model footprint on the embedding, and then reconstructs the raw data by using the inferred model. We show the practicality of the attack on popular DNNs trained for two prominent biometric modalities, face and fingerprint recognition. The attack can effectively infer the original recognition model (mean accuracy 83% for faces, 86% for fingerprints), and can craft effective biometric reconstructions that are successfully authenticated with 1-vs-1 authentication accuracy of up to 92% for some models.

Two parties wish to collaborate on their datasets. However, before they reveal their datasets to each other, the parties want to have the guarantee that the collaboration would be fruitful. We look at this problem from the point of view of machine learning, where one party is promised an improvement on its prediction model by incorporating data from the other party. The parties would only wish to collaborate further if the updated model shows an improvement in accuracy. Before this is ascertained, the two parties would not want to disclose their models and datasets. In this work, we construct an interactive protocol for this problem based on the fully homomorphic encryption scheme over the Torus (TFHE) and label differential privacy, where the underlying machine learning model is a neural network. Label differential privacy is used to ensure that computations are not done entirely in the encrypted domain, which is a significant bottleneck for neural network training according to the current state-of-the-art FHE implementations. We prove the security of our scheme in the universal composability framework assuming honest-but-curious parties, but where one party may not have any expertise in labelling its initial dataset. Experiments show that we can obtain the output, i.e., the accuracy of the updated model, with time many orders of magnitude faster than a protocol using entirely FHE operations.